Dysplasia is the precursor of cancer in a number of gastrointestinal (GI) disorders. Dysplasia is fundamentally a microscopic diagnosis based on specific changes in cells and architecture of the mucosa. A key problem is that dysplastic mucosa is usually not detectable by standard GI endoscopy, a major technical limitation that prevents detection of pre-malignant lesions when therapeutic intervention would be optimal and even curative. Recent studies by us and others have led to the development of new endoscopic approaches that provide an unprecedented quality and accuracy of images of the GI mucosa. This application is focused on the utilization of endoscopic optical coherence tomography (EOCT), a light based imaging modality that generates cross-sectional representations of biological tissues, with a resolution at least ten times greater than that achievable with any of the existing endoscopic imaging modalities. This extremely promising new imaging system is still in its early stages, and intense study is required to solve some of the technological restrictions still existing, such as the small size of the scanning area, imprecise control of the image stream required for microscopic diagnosis, uncertain targeting of the biopsy site, and distortion of the tissue under examination. Based on our preliminary ex vivo and in vivo results, and the integrated efforts of the multidisciplinary team of experts we have assembled, we feel confident that the above limitations can be properly addressed and solved. Thus, the overall goal is to adapt EOCT to the detection of dysplasia in the GI tract and determine its accuracy in the detection of dysplasia in common human pre-malignant conditions. This will be accomplished through the following specific aims: 1) To evaluate prospectively the yield and accuracy of EOCT in endoscopic surveillance of dysplasia in Barrett's esophagus; 2) To use EOCT imaging to measure the density of aberrant colonic crypt foci; and 3) To develop an EOCT probe for imaging pancreatic ductal epithelium, demonstrate the feasibility of pancreatic EOCT, and correlate EOCT images with histological changes of dysplastic ductal epithelium. Achievement of these aims will significantly improve the capacity to detect pre-malignant or early malignant lesions in the GI tract and allow early therapeutic interventions that will positively impact on patient care and survival.